In Vitro Diagnostic Devices in Korea — Why the Medical Device Guides Do Not Apply

Key takeaway — Since 1 May 2020 in vitro diagnostics in Korea have been governed by a separate statute (the In Vitro Diagnostic Devices Act, Act No. 16433). The outline of the process resembles general medical devices, but three things diverge: the basis for classification, the name and structure of the clinical evidence, and the set of notices to consult. Prepare with general device material and the required documents will not line up. Current as of September 2026.

Why it has to be read separately

In vitro diagnostics do not contact the body. They handle specimens and produce information. So the general medical device question — how dangerous is this to the body — does not carry over cleanly.

That difference is what led to a separate regime. The In Vitro Diagnostic Devices Act (Act No. 16433), promulgated on 30 April 2019, took effect on 1 May 2020, giving the field its own statute, enforcement rule and notices.

Why this matters in practice: the medical device material you find by searching may not fit this product. Article numbers, notice titles and forms all differ.

Divergence 1 — the basis for classification

General medical devices are classified by risk to the body. In vitro diagnostics are not.

They are classified into four classes according to the intended use and the potential harm that the resulting diagnostic information could cause to individuals and to public health.

In one sentence: the device does not touch the body, so what sets the class is the harm that incorrect diagnostic information could cause.

The focus of classification therefore moves from the hardware to the intended use. Two analysers working on the same principle can land in different classes depending on what they claim to diagnose, and handling diagnostic information with public health consequences, such as infectious disease, raises the assessment.

The practical implication is clear: the intended-use statement has to be settled before the class emerges. That order is the same as for general devices, but it carries more weight here.

Divergence 2 — the clinical evidence has a different name

Where general medical devices say clinical trial, in vitro diagnostics say clinical performance study.

The structure travels with the name — a clinical performance study plan is prepared and approved by the review committee established at the institution conducting the study. The MFDS maintains separate notices on approval of clinical performance study plans and on management of essential documents for those studies.

The failure mode this creates is simple: different terminology means different regulations. Searching for "medical device clinical trial" leads to rules that do not govern this product. Starting the search under the correct name is more accurate.

Divergence 3 — a dedicated set of notices

There are IVD-specific notices. The ones opened repeatedly in practice:

Notice	What it settles
Regulation on IVD Items and Their Classes	Item classification and class
Regulation on IVD Approval, Notification and Review	Submission documents and review
Regulation on Approval of IVD Clinical Performance Study Plans	Study plan approval
Regulation on Management of IVD Clinical Performance Study Essential Documents	Essential document management

Identically named regulations exist on the general medical device side, so it is worth building the habit of checking which version you opened — look for the IVD prefix in the title.

What carries over — business requirements and the quality system

Company-level requirements share their skeleton with general devices.

• Approval goes to the Minister of Food and Drug Safety; certification or notification goes to the President of the National Institute of Medical Device Safety Information. The channel splits by class
• The quality manager regime applies. The training bodies in fact cover "quality managers for manufacture and import of medical devices and in vitro diagnostics" together — see the quality manager guide
• The structure in which the company-level licence and the product-level item track are separate but parallelisable is also the same — see the manufacturing licence guide

So the summary is: the business and quality layers resemble each other; the content of the product review diverges.

Order of preparation

1. Is this an in vitro diagnostic — does it handle specimens and produce diagnostic information
2. Settle the intended-use wording — what does it claim to diagnose. The class follows from this
3. Fix item classification and class — using the IVD-specific classification regulation
4. Fix the pathway — approval (MFDS) or certification/notification (NIDS)
5. Decide whether a clinical performance study is required — if so, build plan approval into the master schedule
6. Run business requirements in parallel — quality manager, facilities, quality system

Step 5 has the largest schedule impact. Where a performance study is required, its duration sets the whole timeline, so deciding early is what makes planning possible.

Common mistakes

• Applying general medical device guides and forms, producing a document set that does not match
• Searching under "clinical trial" and reading rules that do not govern this product
• Guessing the class from the hardware, without reflecting the risk created by the intended use
• Underweighting diagnostic information with public health consequences, such as infectious disease
• Deciding late whether a clinical performance study is needed, pushing the whole schedule back

Before you start

• Confirm the product is an in vitro diagnostic device
• Draft the intended-use wording — what is diagnosed, and for what purpose
• Confirm item and class using the IVD classification regulation (do not mix it with the general device one)
• Fix the pathway — approval (MFDS) or certification/notification (NIDS)
• Decide whether a clinical performance study is required → secure plan approval lead time
• Design quality manager and facility requirements in parallel with the item track

---

For in vitro diagnostics, the intended-use wording sets the class, and the class sets everything else. Tell us the specimen type and the diagnostic purpose you intend to claim, and we will identify the likely class range and whether a clinical performance study is in the picture — see free preliminary review.